Annals of Surgical Oncology

, Volume 16, Issue 1, pp 140–146

Pain Palliation in Patients with Bone Metastases Using MR-Guided Focused Ultrasound Surgery: A Multicenter Study

Authors

    • Division of OrthopedicsSheba Medical Center
  • David Gianfelice
    • Department of Diagnostic ImagingToronto General Hospital
  • Yael Inbar
    • Department of Diagnostic ImagingSheba Medical Center
  • Alexander Beck
    • Department of RadiologyCharite Hospital
  • Tatiana Rabin
    • Oncology InstituteSheba Medical Center
  • Noga Shabshin
    • Department of Diagnostic ImagingSheba Medical Center
  • Gupta Chander
    • Department of Diagnostic ImagingToronto General Hospital
  • Suzanne Hengst
    • Department of RadiologyCharite Hospital
  • Raphael Pfeffer
    • Oncology InstituteSheba Medical Center
  • Aharon Chechick
    • Division of OrthopedicsSheba Medical Center
  • Arik Hanannel
    • InSightec Ltd.
  • Osnat Dogadkin
    • InSightec Ltd.
  • Raphael Catane
    • Oncology InstituteSheba Medical Center
Palliative Care

DOI: 10.1245/s10434-008-0011-2

Cite this article as:
Liberman, B., Gianfelice, D., Inbar, Y. et al. Ann Surg Oncol (2009) 16: 140. doi:10.1245/s10434-008-0011-2

Abstract

Background

Noninvasive thermal ablation using magnetic resonance (MR)-guided focused ultrasound (MRgFUS) has been shown to be clinically effective in uterine fibroids, and is being evaluated for ablation of breast, liver, and brain lesions. Recently MRgFUS has been evaluated for palliation of pain caused by bone metastases. We present the clinical results of a multicenter study using MRgFUS for palliation of bone metastases pain.

Methods

A multicenter study to evaluate the safety and efficacy of MRgFUS palliative treatment of bone metastases was conducted in patients suffering from painful metastatic bone lesions for which other treatments were either ineffective or not feasible.

Thirty-one patients with painful bone metastases underwent the MRgFUS procedure in three medical centers. Treatment safety was evaluated by assessing the device-related complications. Effectiveness of pain palliation was evaluated using the visual analog pain score (VAS), and measurable changes in the intake of opioid analgesics.

Results

Thirty-six procedures were performed on 31 patients. Mean follow-up time was 4 months. 25 patients underwent the planned treatment and were available for 3 months post-treatment follow-up. 72% of the patients (18/25) reported significant pain improvement. Average VAS score was reduced from 5.9 prior to treatment to 1.8 at 3 months post treatment. 67% of patients with recorded medication data reported a reduction in their opioid usage. No device-related severe adverse events were recorded.

Conclusion

The results suggest that MRgFUS has the ability to provide an accurate, effective, and safe noninvasive palliative treatment for patients with bone metastases.

Bone pain and mechanical bone impairment due to skeletal lesions are common among patients suffering from cancer metastases. Among breast cancer patients 90% of patients dying from this disease suffer from bone metastases,13 and most of the patients with metastatic prostate cancer have active bone lesions.4,5 Thirty percent of all cancer patients will develop bone metastases6 and 50–57% of these suffer from significant pain.7 The increasing longevity of cancer patients coupled with a rise in cancer incidence over the last decade have contributed profoundly to the growing number of patients with bone metastases. Better cancer treatments and increased emphasis on the quality of life of cancer patients has led to a search for effective pain palliation for clinically active bone metastases, with fewer short- and long-term side-effects.

Bone pain is the most common source of pain in patients suffering from metastatic cancer.1,8 Current treatments for this situation include systemic therapies such as analgesics, chemotherapy, hormonal therapy, biphosphonates, and local control such as radiation therapy,911 radiofrequency ablation,1214 and surgical stabilization of the affected site.

External beam radiation is the current standard of care for clinically active bone metastases. However, 20–30% of the patients do not experience any pain relief,3 and in 27% pain recurs;15,16 however, they are no longer eligible for radiation treatments due to the dose load accumulated from previous treatments.

Magnetic resonance (MR)-guided focused ultrasound (MRgFUS) is a noninvasive thermal ablation technique that enables the physician to perform localized ablation of tumor tissue, by focusing the acoustic energy precisely at the targeted volume. MR guidance allows real-time monitoring of location and intensity of temperature rise at the treatment site.

The concept of focused ultrasound as a noninvasive treatment alternative was introduced by Lele over 40 years ago.17 Jolesz and Hynynen presented the concept of MR-guided focused ultrasound in 1992/1993,18,19 and since then it has been shown to be clinically effective in treating uterine fibroids,2023 and is currently being evaluated for treatment of breast,2426 liver,27 and brain tumors.28,29

Treating bone lesions with MRgFUS, in contrast to treating soft tissue tumors (STT), has certain advantages, such as higher absorption of the acoustic energy (up to 50 times higher than STT absorption) and low thermal conductivity of bone tissue.18,30 These differences lead to limited penetration into normal cortical bone when using the soft tissue tumor MRgFUS system. Applying acoustic energy on the bone surface allows the creation of temperature rise in the part of the bone cortex enclosed in the beam path zone, thus indirectly ablating the adjacent periosteum and tumor tissue. Since the bone periosteum is considered to be a major source of pain in patients with metastatic bone lesions, ablating the source of pain should produce lasting pain relief.2 Recently MRgFUS was evaluated for palliation of pain caused by bone metastases with very promising initial results.31 The aim of this study was to continue assessment of the safety and efficacy of MRgFUS in treating painful bone metastases.

Patients and Methods

Thirty-one patients with painful bone metastasis were treated with MRgFUS surgery in three medical centers: the Sheba Medical Center (Tel Hashomer, Tel-Aviv, Israel), Toronto General Hospital (Toronto, Ontario, Canada), and Charite Hospital (Berlin, Germany). The study was conducted in full accordance with the local internal review boards, and the national Helsinki committee approval and guidelines.

The main inclusion criteria were clinically identifiable painful bone metastatic lesions, in patients who had exhausted or refused all other pain palliation methods including external beam radiation. MRgFUS treatment was conducted using a focused ultrasound phased array system (ExAblate® 2000 system, InSightec Ltd., Haifa, Israel), integrated with a magnetic resonance imaging (MRI) scanner (GE 1.5-T MRI, Milwaukee, WI, USA).

The procedure was performed with the patient lying inside the MRI scanner, under conscious sedation and analgesia using 5–10 mg IV midazolam and 2–40 mg IV morphine sulfate, administrated by the oncologist, depending on the previous patient’s levels of opioid treatment. One patient underwent regional spinal anesthesia due to intractable pretreatment positional pain.

Based on pretreatment MRI and computed tomography (CT) screening images, which were used to identify the treatment region, the patient was placed on the patient table with the targeted lesion positioned over a water bath containing the ultrasound transducer (Fig. 1).
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-008-0011-2/MediaObjects/10434_2008_11_Fig1_HTML.jpg
Fig. 1

MRgFUS treatment: a schematic display. Patient is lying supine on the ExAblate® 2000 patient table, the targeted lesion is positioned over a water bath containing the ultrasound-generating transducer and robotic arm (positioning system) that provides movement of the transducer from one treatment point to another. Ablated area is produced by intersection of acoustic beam (blue) and bony tissue, where practically all of the energy is absorbed (hence the tip of the blue cone is outlined only).

Acoustic coupling was accomplished by placing a coupling gel pad between the patient and patient table.

The position of the patient and the sonication pathway were checked using MR imaging (standard T2-weighted fast spin-echo images). Images were then loaded into the MRgFUS workstation and the target area and bone contour were delineated for treatment. Following this the workstation generated a patient-specific (personalized) treatment plan that avoided damage to nontarget tissue while optimizing the required energy level and the number of sonications.

At the beginning of each procedure a number of low-power sonications were performed to ensure the targeting accuracy in three dimensions. Treatment at therapeutic power levels began after a mild increase in the temperature at the expected position was detected.

Throughout the treatment, each sonication’s location and the temperature elevation in the tissue adjacent to the targeted bone were monitored in real time, using the proton resonance frequency (PRF) shift temperature measurement method. In response to the resulting temperature map the treating physician could decide to modify treatment parameters such as power, frequency, sonication duration or spot size.

At the end of each procedure, contrast-enhanced MRI scans (fat-suppressed T1-weighted contrast-enhanced spoiled gradient-recalled-echo images) were performed to ensure that the extent of ablation was confined to target tissue and that there was no substantial damage in the tissue surrounding the target.

Patients were followed up at 3 days, 2 weeks, and 1 and 3 months after treatment. One site (Charite) followed up patients at 6 months as well. Each follow-up visit included a full physical examination, visual analog pain score questionnaire (VAS),32 and medication level questionnaires, providing data on direct and indirect changes in pain levels. The 3-month visit included an MR scan (T2 and T1 weighted, with and without contrast) and CT imaging.

Treatment response was categorized using the endpoint criteria defined by the International Bone Metastases Working Party guidelines on palliative radiotherapy endpoints for future clinical trials,33 according to which partial response is defined as a drop of 2 points in the VAS score without an increase in pain medications or a drop of 25% in pain medication without increase in reported pain score. Complete response is defined as a pain score of 0 on the VAS score without medication increase (Table 1). The evaluation of changes in medications was calculated using morphine equianalgesic doses (Fig. 2).
Table 1

Criteria for treatment response as defined by the international consensus on palliative radiotherapy paper33

Criteria for pain palliation treatment outcome

Partial response

Pain reduction of 2 or more at the treated site on a 0–10 scale without analgesic increase

Analgesic reduction of 25% or more from baseline without an increase in pain

Complete response

Pain score of zero at the treated site with no concomitant increase in analgesic intake

https://static-content.springer.com/image/art%3A10.1245%2Fs10434-008-0011-2/MediaObjects/10434_2008_11_Fig2_HTML.gif
Fig. 2

Average VAS score over time.

Treatment safety was monitored by recording the incidence of any device-related morbidity, either local or systemic at each follow-up point.

Results

Patient Population

Average patient age was 61 years (range 40–85 years) with an almost equal gender distribution (15 men and 16 women). The most common treatment site was iliac bone (18/31), and the most common primary tumor type was breast cancer (11/31). The ratio between osteoblastic and osteolytic tumors was 1:2. Table 2 lists patients’ demographic data and tumor characteristics.
Table 2

Patient characteristics

Patient characteristics

Number of patients treated

31

Female

16

Male

15

Age, median [range], years

61 [40–85]

Primary tumor type

Renal CA

6

Colorectal CA

2

Lung CA

1

Breast CA

11

Prostate CA

5

Other

6

Targeted lesion type

Osteolytic

20

Osteoblastic

10

Prior radiation to treated site

21/31

Concurrent opioid analgesics

10*

Targeted lesion location

Iliac bone

18

Ischium bone

4

Sacrum

4

Femur

1

Scapula

2

Humerus

1

Clavicle

1

* Excluding patients lost to follow-up.

Thirty-six treatments in 31 patients were conducted, targeting 32 metastatic lesions. The average length of the treatment was 66 min (range 22–162 min). The average number of sonications was 17.3 (range 8–32), with average sonication energy of 1,135 J (range 440–1,890 J). Mean follow-up time was 108 days after treatment.

No device-related severe adverse events were recorded in any of the patients. All patients except two tolerated the MRgFUS treatment using only conscious sedation. The general health of one patient deteriorated in the days after the treatment leading to his death. A second patient died before reaching the 3-month follow-up examination due to the advanced stage of his disease. Another patient was unable to tolerate the treatment due to pain aggravation before starting the treatment and was excluded from the study.

An efficacy VAS analysis was conducted on 25 patients who had full treatment and reached the 3-month follow-up visit. The mean baseline VAS score before treatment was 5.9 (range 3.5–8.5), mean VAS score 3 days after treatment was 3.8 (range 0–8.5), and by the 3-month post-treatment follow-up it had dropped to 1.8 (range 0–8) (Fig. 3). A reduction of two points on VAS scale at 3 months proved to be statistically significant (P < 0.003).
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-008-0011-2/MediaObjects/10434_2008_11_Fig3_HTML.gif
Fig. 3

Patients’ response to treatment at 3-month follow-up.

Seventy-two percent of patients (18/25) had a significant reduction in pain (>2 points) at the 3-month follow-up. Of these 50% (9/25) reported a VAS score of 0. Twenty-four percent had no response and one patient (4%) experienced worsened pain levels (Fig. 4).
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-008-0011-2/MediaObjects/10434_2008_11_Fig4_HTML.jpg
Fig. 4

Targeted osteolytic lesion, located in the superior part of right iliac bone, as seen on pre treatment T2-weighted axial MR image (left) and immediate post-treatment T1-weighted contrast-enhanced fat sat axial MR image (right). A small nonenhancing area can be seen in the treated region (arrow), surrounded by hyperemia induced by soft tissue edema.

It is significant to point out that 52% of patients reported substantial pain relief as early as at the 3 days post treatment (Fig. 3).

Twelve out of 25 patients used opioid analgesics and we were able to collect medication data on 10 of these patients. Sixty seven percent of these decreased their use of analgesic medication, and 22% had increased their medication dosage (Table 3). Most of the patients taking nonopioid analgesics (13/25) also reported a drop in medication usage.
Table 3

Medications – only the data of patients who used concurrent opioid analgesics is displayed, excluding patients lost to follow-up

 

Medications, morphine equianalgesic dose [mg]

Pt. #

Baseline

3 days

2 weeks

1 month

3 months

Change from baseline

Change %

4

40

40

40

40

40

0

0

8

200

200

160

160

160

40

20

16

209

315

305

345

345

−136

−65

18*

520

200

200

0

0

520

100

19

90

120

120

520

150

−60

−67

20

180

60

60

88

140

40

22

21

55

45

45

45

45

10

18

23

10

10

0

0

0

10

100

24

120

120

100

100

100

20

17

29

30

30

30

30

30

0

0

* This patient was excluded from medications analysis since opioid analgesic was changed to none morphine equianalgesic intake.

When combining both the medication and VAS score, 36% of patient had partial response and 36% had complete response according to the working group criteria.33

We found no correlation between the response to the treatment and patient gender or lesion type (osteolytic versus osteoblastic). Immediate post-treatment contrast-enhanced MR images showed edema at the target area and in some cases minor localized ablation (Fig. 5). MR images at 3 months did not show evidence of lasting damage, and follow-up CT images showed some cases of calcification of targeted tissue (Fig. 5).
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-008-0011-2/MediaObjects/10434_2008_11_Fig5_HTML.jpg
Fig. 5

Pre-treatment (upper row) and 3-months post-treatment (lower row) CT images of treated lesion in right iliac bone. Evidence of new bone formation is seen in the treated area (encircled).

Discussion

Pain palliation is one of the most important factors influencing the quality of life in patients with metastatic bone lesions.15,8 MRgFUS treatment has the potential of avoiding the side-effects caused by other therapies, by using noninvasive, nonionizing, accurate, controlled ablation.3438

It is possible that the rapid improvement of pain is due to local bone denervation, caused by the heat denaturation of the periosteum layer in the treated area. Since the energy is nonionizing, there is no upper limit to the accumulated acoustic energy allowed to pass through adjacent soft tissue as long as tissue temperature is kept at the safe level. Thus the treatment can be repeated as needed, without reaching maximal dose as in the case of external beam radiation therapy.

Due to the low thermal conductivity and high absorption rate of the bone cortex it is possible to use lower energy levels for this treatment compared with when using soft tissue treatments, and improve the safety profile by reducing the thermal damage around the treated bone site. The temperature monitoring during the treatment was limited to the soft tissue adjacent to targeted bone for two reasons: (1) due to the low water content of bone cortex it is practically impossible to calculate the temperature in that area, (2) the PRF method cannot be applied in fat tissue, thus temperature measurements in bone marrow are unreliable. Nonetheless, the received thermal feedback was sufficient to allow control over actual sonication location and minimization of damage to nontarget tissues.

In this multicenter study the mean VAS score dropped from 5.9 at baseline to 1.8 after 3 months. Even considering the relatively few patients in this study, VAS reduction was statistically significant for a 2 points change (P < 0.003).

Although most of the patients had exhausted their nonmedication palliative alternatives, we observed a favorable response in 72% of patients treated. It is also important to note that most patients reported substantial pain relief as early as 3 days after treatment, that only one session was required in most cases, and that treatment durability seems to extend to at least 3 months. Moreover MRgFUS was equally effective in achieving pain palliation in both osteoblastic and osteolytic bone metastases. No device-related serious adverse events occurred in any of the patients that were treated.

Although the study protocol was designed to include only patients with localized, well-defined painful lesions, the patients recruited into this initial feasibility study were by nature and definition very ill and in some cases could not distinguish the pain arising from a specific intended-to-treat lesion. When pain is not localized, it may be difficult to identify the origin of pain, especially where there are several metastases. Although every effort was made to obtain pain assessments specific to the site where the patient was treated, in some cases this was not possible and total bodily pain was reported instead. We suspect that the presence of such confusing data could be responsible for the 28% rate of treatment failures, since no other significant differences were found in these cases.

Depending upon the location of the lesion, there were occasional difficulties in positioning the patient over the water bath with the ultrasound-generating transducer.

The current system design limits to some extent the anatomical locations suitable for the treatment. In the future, a more flexible transducer assembly should be considered, to expand significantly the indications for treatment.

Since this was not a randomized controlled study, it is impossible to decide objectively whether this technique is superior to other existing palliation modalities. The purpose of this study was to introduce a novel use of MR-guided focused ultrasound for pain palliation of bone metastasis. Although, it is generally very difficult to perform prospective randomized controlled studies with an interventional medical device, especially so in metastatic cancer patients, future studies will have to address the issue of the value of MRgFUS and other palliation techniques.

Issues that will need to be answered in future studies include durability over a longer follow-up period, the use of larger patient population, and correlation between site of metastases or primary tumor type and treatment outcome. Another interesting area to explore is the potential synergy between MRgFUS and other cancer therapies, such as chemotherapy or radiation.

We have seen anecdotal evidence that treated lesion was affected, but larger studies would be needed to understand better and maybe even predict the extent of such an effect.

In summary, the above results show that the MRgFUS technology has the potential to be an important tool in treating painful bone metastases.

Copyright information

© Society of Surgical Oncology 2008